Optimal QoS aware multiple paths web service composition using heuristic algorithms and data mining techniques
The goal of QoS-aware service composition is to generate optimal composite services that satisfy the QoS requirements defined by clients. However, when compositions contain more than one execution path (i.e., multiple path's compositions), it is difficult to generate a composite service that si...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | eng eng |
| Published: |
2014
|
| Subjects: | |
| Online Access: | https://etd.uum.edu.my/5314/1/s91962.pdf https://etd.uum.edu.my/5314/2/s91962_abstract.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The goal of QoS-aware service composition is to generate optimal composite services that satisfy the QoS requirements defined by clients. However, when compositions contain more than one execution path (i.e., multiple path's compositions), it is difficult to generate a composite service that simultaneously
optimizes all the execution paths involved in the composite service at the same time while meeting the QoS requirements. This issue brings us to the challenge of solving the QoS-aware service composition problem, so called an optimization problem. A further research challenge is the determination of the QoS characteristics that can be considered as selection criteria. In this thesis, a smart QoS-aware service composition approach is proposed. The aim is to solve the above-mentioned problems via an optimization mechanism based upon the combination between runtime path prediction method and heuristic algorithms. This mechanism is performed in two steps. First, the runtime path prediction method predicts, at runtime, and just before the actual composition, execution, the execution path that will potentially be executed. Second, both the constructive procedure (CP) and the complementary procedure (CCP) heuristic algorithms computed the optimization considering only the execution path that has been predicted by the runtime path
prediction method for criteria selection, eight QoS characteristics are suggested after
investigating related works on the area of web service and web service composition. Furthermore, prioritizing the selected QoS criteria is suggested in order to assist clients when choosing the right criteria. Experiments via WEKA tool and simulation prototype were conducted to evaluate the methods used. For the runtime path prediction method, the results showed that the path prediction method achieved promising prediction accuracy, and the number of paths involved in the prediction did not affect the accuracy. For the optimization mechanism, the evaluation was conducted by comparing the mechanism with relevant optimization techniques. The simulation results showed that the proposed optimization mechanism outperforms the relevant optimization techniques by (1) generating the highest overall QoS ratio solutions, (2) consuming the smallest computation time, and (3) producing the lowest percentage of constraints violated number. |
|---|